{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9893390191897654,
  "eval_steps": 100,
  "global_step": 58,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "clip_ratio": 0.0,
      "completion_length": 603.7678756713867,
      "epoch": 0.017057569296375266,
      "grad_norm": 0.5337616801261902,
      "kl": 0.0,
      "learning_rate": 5e-07,
      "loss": 0.0241,
      "reward": 0.628348246216774,
      "reward_std": 0.34275937266647816,
      "rewards/accuracy_reward": 0.628348246216774,
      "rewards/format_reward": 0.0,
      "step": 1
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 597.266489982605,
      "epoch": 0.08528784648187633,
      "grad_norm": 0.3713057339191437,
      "kl": 0.00017765164375305176,
      "learning_rate": 2.5e-06,
      "loss": 0.0004,
      "reward": 0.6060268161818385,
      "reward_std": 0.36045964574441314,
      "rewards/accuracy_reward": 0.6060268161818385,
      "rewards/format_reward": 0.0,
      "step": 5
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 615.9348472595215,
      "epoch": 0.17057569296375266,
      "grad_norm": 34.452064514160156,
      "kl": 0.007735443115234375,
      "learning_rate": 2.956412726139078e-06,
      "loss": 0.0672,
      "reward": 0.692410746216774,
      "reward_std": 0.3009307190775871,
      "rewards/accuracy_reward": 0.6921875312924385,
      "rewards/format_reward": 0.00022321429569274187,
      "step": 10
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 599.2725708007813,
      "epoch": 0.255863539445629,
      "grad_norm": 0.12593211233615875,
      "kl": 0.006564712524414063,
      "learning_rate": 2.7836719084521715e-06,
      "loss": 0.0718,
      "reward": 0.764062537252903,
      "reward_std": 0.21816119737923145,
      "rewards/accuracy_reward": 0.764062537252903,
      "rewards/format_reward": 0.0,
      "step": 15
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 587.9596244812012,
      "epoch": 0.3411513859275053,
      "grad_norm": 23.166851043701172,
      "kl": 0.01224517822265625,
      "learning_rate": 2.4946839873611927e-06,
      "loss": 0.0505,
      "reward": 0.7772321745753288,
      "reward_std": 0.1896885072812438,
      "rewards/accuracy_reward": 0.7772321745753288,
      "rewards/format_reward": 0.0,
      "step": 20
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 593.9265853881836,
      "epoch": 0.42643923240938164,
      "grad_norm": 0.114225834608078,
      "kl": 0.004923629760742188,
      "learning_rate": 2.1156192081791355e-06,
      "loss": 0.044,
      "reward": 0.7667411044239998,
      "reward_std": 0.1819651648402214,
      "rewards/accuracy_reward": 0.7667411044239998,
      "rewards/format_reward": 0.0,
      "step": 25
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 598.4111846923828,
      "epoch": 0.511727078891258,
      "grad_norm": 0.18414755165576935,
      "kl": 0.005031585693359375,
      "learning_rate": 1.6808050203829845e-06,
      "loss": 0.0343,
      "reward": 0.7582589626312256,
      "reward_std": 0.1700172070413828,
      "rewards/accuracy_reward": 0.7582589626312256,
      "rewards/format_reward": 0.0,
      "step": 30
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 587.543773651123,
      "epoch": 0.5970149253731343,
      "grad_norm": 1.1138083934783936,
      "kl": 0.00536346435546875,
      "learning_rate": 1.2296174432791415e-06,
      "loss": 0.033,
      "reward": 0.7468750327825546,
      "reward_std": 0.175389445386827,
      "rewards/accuracy_reward": 0.7468750327825546,
      "rewards/format_reward": 0.0,
      "step": 35
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 579.6279319763183,
      "epoch": 0.6823027718550106,
      "grad_norm": 0.11763972043991089,
      "kl": 0.004239654541015625,
      "learning_rate": 8.029152419343472e-07,
      "loss": 0.0394,
      "reward": 0.7729911059141159,
      "reward_std": 0.1725513393059373,
      "rewards/accuracy_reward": 0.7729911059141159,
      "rewards/format_reward": 0.0,
      "step": 40
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 591.583285522461,
      "epoch": 0.767590618336887,
      "grad_norm": 0.15979616343975067,
      "kl": 0.005073165893554688,
      "learning_rate": 4.3933982822017883e-07,
      "loss": 0.0356,
      "reward": 0.7535714641213417,
      "reward_std": 0.1820222762878984,
      "rewards/accuracy_reward": 0.7535714641213417,
      "rewards/format_reward": 0.0,
      "step": 45
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 596.1100730895996,
      "epoch": 0.8528784648187633,
      "grad_norm": 0.32247623801231384,
      "kl": 0.005157852172851562,
      "learning_rate": 1.718159615201853e-07,
      "loss": 0.0349,
      "reward": 0.742857177555561,
      "reward_std": 0.19252488762140274,
      "rewards/accuracy_reward": 0.742857177555561,
      "rewards/format_reward": 0.0,
      "step": 50
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 581.0707809448243,
      "epoch": 0.9381663113006397,
      "grad_norm": 0.1405262053012848,
      "kl": 0.013854598999023438,
      "learning_rate": 2.4570139579284723e-08,
      "loss": 0.0351,
      "reward": 0.7852678924798966,
      "reward_std": 0.18969743903726338,
      "rewards/accuracy_reward": 0.7852678924798966,
      "rewards/format_reward": 0.0,
      "step": 55
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 581.581911722819,
      "epoch": 0.9893390191897654,
      "kl": 0.0072021484375,
      "reward": 0.7686012263099352,
      "reward_std": 0.1908461432904005,
      "rewards/accuracy_reward": 0.7686012263099352,
      "rewards/format_reward": 0.0,
      "step": 58,
      "total_flos": 0.0,
      "train_loss": 0.03941735033154616,
      "train_runtime": 7213.9356,
      "train_samples_per_second": 1.04,
      "train_steps_per_second": 0.008
    }
  ],
  "logging_steps": 5,
  "max_steps": 58,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": false,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}